Embryonic stem cells (ESCs) are envisioned as a source of transplantable pancreatic islets to alleviate the donor organ shortage. Despite many recent advances, protocols to differentiate ESCs to homogeneous pools of glucose-responsive insulin-secreting beta cells have yet to be developed. It is widely accepted that the production of these populations will first require generation of definitive endoderm. Thus, the successful derivation and stable propagation of endoderm-restricted stem cell lines would be a significant step forward in the development of pancreas, liver, and lung cells for cell replacement therapies and for improving our understanding of how these tissues are normally formed from this germ layer. We have developed a simple protocol, using magnetic-activated cell sorting (MACS), to isolate endoderm-committed cells from heterogeneous cultures of differentiated murine ESCs on the basis of epithelial cell adhesion molecule (EpCAM) expression. The protocol includes negative selection to remove undifferentiated ESCs and visceral endoderm cells, and positive selection to retain cells that express EpCAM. When sorted cells are placed sub-cutaneously into immunodeficient mice, small EpCAM+PDX1+ nodules highly reminiscent of embryonic pancreatic epithelium develop. Cells from these nodules can be extensively propagated in vitro and have the characteristics expected of an endodermal stem cell (EndSC) or foregut- restricted stem cell (FGSC) population, depending on the duration of growth in vivo. The EndSCs are OCT4+ and express some early endoderm-restricted genes, whereas FGSCs are OCT4- and have high expression of Sox17, Foxa2, and PDX1. EndSCs and FGSCs transplanted back into animals recapitulate the original nodule phenotype, demonstrating their stability in vitro. We hypothesize that only the OCT4+EpCAM+SSEA1- cells recovered from the MACS separation give rise to the in vivo-derived endodermal cell lines. Here, we will directly test this hypothesis by FACS-sorting OCT4+EpCAM+SSEA1- cells using an OCT4GFP ESC line. Cell lines derived from the FACS-sorted population will be compared to the existing MACS-derived EndSCs and FGSCs. We will also study the ability of EndSCs and FGSCs to differentiate in vitro to specific endoderm-derived cell types, including beta cells. To promote differentiation, we will culture cells in an embryonic pancreas environment, expose cells to relevant growth factor signals, and culture cells according to recently published protocols. These studies will contribute not only to improved differentiation efficiency of ESCs to definitive endoderm and pancreatic lineages, but also to a better understanding of germ layer segregation and early embryonic cell fate decisions. PUBLIC HEALTH RELEVANCE: Our preliminary studies have led to the development of cell lines from differentiated mouse embryonic stem cells (ESCs) that appear to be foregut endoderm-committed. These cell lines are non-tumorigenic and grow well in culture, and here we propose to elucidate the cell of origin of these endodermal cell lines and explore their ability to differentiate into pancreatic, liver, and lung cells in vitro. This knowledge will help in establishing protocols for the isolation and differentiation of human endoderm-committed cells for the production of functional cells to be used in transplantation therapies.